This is a sample of 543 documented cases of child neglect followed 30 years later. All cases were less than 11 years of age at the time the neglect/ abuse was documented. Neglect cases will be compared to cares of documented sexual abuse (N= 96) and documented physical abuse (N = 110) and a matched comparison group (N = 520). Assessments will include: (a) health status and health risk behavior; (b) economic productivity; (c) neighborhood hazards/ toxins; and (d) past and present service utilization and access to care. This study will also develop a self-report measure of neglect and assess correspondence of memories of neglect with documented history of neglect. Similar measures were developed in past research with this sample with regard to self report of childhood sexual and physical abuse. Some preliminary analyses from previous studies of this population have documented relatively high levels of health concerns, exposure to subsequent traumatic life events (e.g., rape) leading to PTSD, and relatively low SES. Within each domain of inquiry, several specific hypotheses are noted. Measures are drawn from previous epidemiological survey research, including the National Health Interview Survey, the Behavioral Risk Factor Surveillance Survey, and the Social Support measure from the Vietnam Veterans Readjustment Study. Some established questionnaires will be administered to participants to measures self esteem, locus of control, stressful life events, coping, and depression. Some limited neurobehavioral assessments will be completed, as well. Neighborhood characteristics will be assessed from census data on poverty, as well as by asking participants to describe their neighborhood during an interview. Economic productivity will be assessed using information about income, public assistance, size of home, etc. Plans are in place for a medical screening that will include immunological functioning and HIV status. The analysis plans were detailed and appropriate. Group comparisons will be made using MANOVAs with covariance, as necessary. Power analyses indicate adequate sample size, especially since the CSA and CPA groups will be aggregated for many of the analyses. Hierarchical multiple regressions are planned, as well as hierarchical linear modeling. Strategies for dealing with missing data are described. The investigators make appropriate cautions against over generalizing the results since current definitions and practices for documenting child abuse and neglect differ from those used 30 years ago.